MetaTaskAlgorithm<T, TInput, TOutput>

Models & Types · AiDotNet.MetaLearning.Algorithms

Implementation of MetaMix (Chen et al., ICPR 2020): MAML whose outer loss is computed on virtual examples mixed from pairs of query examples.

Inheritance: ModelBase<T, TInput, TOutput> → MetaLearnerBase<T, TInput, TOutput> → MetaTaskAlgorithm<T, TInput, TOutput>

How It Works

Algorithm (paper Algorithm 1, Eq. 2-7):

MixSource can instead draw the first example of each pair from the support set. Inputs and targets must be Tensor, Matrix or Vector with examples along the first axis.

Constructors

Constructor Summary
MetaTaskAlgorithm(MetaTaskOptions<T, TInput, TOutput> options)

Properties

Property Summary
MetaLearningAlgorithmType AlgorithmType { get; }

Methods

Method Summary
IModel<TInput, TOutput, ModelMetadata<T>> Adapt(IMetaLearningTask<T, TInput, TOutput> task)
T MetaTrain(TaskBatch<T, TInput, TOutput> taskBatch)